The digitalis glycosides, which are so well known because they have been widely used clinically for more than two centuries, are representatives of a much larger group of potent and specific steroidal inhibitors of Na,K-ATPase. The Na,K-ATPase inhibitory moieties of these compounds all belong to one of two structurally related, yet distinct, families of compounds: cardenolides and bufadienolides. Cardenolides and bufadienolides both possess similar steroidal structures, which are somewhat unusual among naturally occurring compounds in that their C and D rings are oriented in a cis configuration. Cardenolides differ structurally from bufadienolides only in that the lactone ring linked to the C17 position on the D ring of cardenolides is a five-membered .gamma. lactone ring, while the lactone ring of bufadienolides is a six-membered .delta. lactone ring. The importance of the lactone ring is underscored by the fact that unconjugated bufadienolides with .delta. lactone rings are approximately tenfold more potent as Na,K-ATPase inhibitors than unconjugated cardenolides with identical steroidal structures, but with .gamma. lactone rings.
Both cardenolides and bufadienolides found widely distributed throughout the plant and animal kingdoms and are almost invariably in the form of C3 conjugates, principally glycosides and esters. Toad conjugates differ from plant conjugates in that they are conjugated to arginyl hemisuberate residues rather than sugar residues. A very striking feature of all of these conjugates is the fact that, in the species in which they are most abundant (Digitalis, Scilla [squill], dozens of other plants, certain insects, bufonid toads, and Rhabdophis tigrinus, a snake), the steroidal molecules and their conjugates are quite heterogeneous, with dozens of molecular species often present. Plant conjugates (of both bufadienolides and cardenolides) are glycosides, while toad conjugates most commonly are carboxyl esters. The best characterized toad carboxyl esters are the arginyl suberate esters of toad bufadienolides, which are abundant in toad skin and cutaneous glandular secretions (11), and which are known as bufotoxins; small amounts of similar arginyl suberate esters of cardenolides ("cardenobufotoxins") have also been described in toad skin and cutaneous secretions.
Unconjugated bufadienolides and cardenolides are sometimes referred to as genins or, alternatively, when derived from plant glycosides, they may be referred to as aglycones. The active steroid components of bufotoxins are also referred to as genins. The cardiac genins are the active moieties of cardiac glycosides. The cardiac glycosides are a large group of sugar conjugates of cardiotonic steroids derived from a variety of plant sources including Digitalis, Strophanthus, Convallaria, Scilla and Helleborus. Except for the unconjugated bufadienolides which are often present in appreciable amounts in toad cutaneous glandular secretions, genins are rarely present as major components of fresh specimens obtained from plants, toads, or insects. Toad bufotoxins and plant bufadienolide glycosides are usually comparable in Na,K-ATPase inhibitory potency with corresponding bufadienolide genins; on the other hand, cardenolide glycosides are generally at least one order of magnitude more effective as Na,K-ATPase inhibitors than are their corresponding cardenolide genins.
The neural effects of digitalis and related compounds have been appreciated for most of this century and there is considerable recent evidence to suggest that many of the beneficial effects of digitalis in congestive heart failure are neurally mediated. It is also known that ouabain, injected into a nerve can cause nerve block (2) and that several bufadienolide genins are effective topical anesthetics (1,9,12,16). However, there have been relatively few experimental efforts to explore the potential therapeutic neural effects of systemic digitalis administration in noncardiac disease. It has been generally thought that, because of the relatively poor penetrance of the blood-brain barrier (BBB) by clinically available glycosides, therapeutically acceptable doses of these glycosides would not exert clinically significant neural effects. Cardenolide and bufadienolide genins, which readily cross the BBB, have rarely been used clinically, even though their neural-to-myocardial tissue concentration ratio is such that one can theoretically achieve a five-to ten-fold increase in CNS levels while halving the myocardial concentration. Reluctance to employ genins presumably stems from the fact that the aglycones (genins) of the clinically used digitalis glycosides are only about one tenth as potent as the corresponding glycosides. Since, until recently, most physicians believed that much of the beneficial clinical effect of digitalis stemmed from its direct myocardial actions, there was an understandable belief that less potent, but potentially neurotoxic agents, would have limited, it any, clinical utility.
The action of cardiac glycosides and genins on nerve conduction has been studied in experimental animals. Ouabain, injected into a nerve, can cause nerve block (2). Several bufadienolide genins have been shown to be effective as topical anesthetics (1,9,12,16).
Additionally, during the past decade, several reports have appeared, providing evidence that, by virtue of its ability to inhibit Na,K-ATPase, ouabain, an especially polar glycoside, can enhance neural conduction in several experimental models of demyelination (4,5,6,10,14), presumably by lowering the resting potential of hyperpolarized, demyelinated neurons (via inhibition of the sodium-potassium pump, or Na,K-ATPase). The less polar cardenolide glycoside, digoxin, has been given to seven patients with multiple sclerosis and to three patients with Pelizaeus-Merzbacher disease. Slight improvement in symptoms and in neurophysiological measurements were noted in six of the ten patients (3,8). It was noted that the effects were limited by dosage constraints (related to the risk of cardiac arrhythmias) and by limited passage of digoxin across the blood brain barrier (3,8,7).
Both the bufadienolide glycosides and the cardenolide glycosides (mainly, the latter) have been used extensively for more than sixty years in the treatment of congestive heart failure and certain cardiac arrhythmias.
There are no records of plant bufadienolide genins being used clinically, or of cardenolide genins used therapeutically in humans, nor of cardiac genins used in treatment of neurological disease. A toad bufadienolide genin (cinobufagin) was administered to patients with atrial fibrillation. Intravenous doses of 1 mg rapidly slowed the heart rate of some patients, but the effect was dissipated within five hours (indicating a much shorter time of action than digoxin or digitoxin, which are administered only once daily). Doses of 2-4 mg were toxic (nausea and vomiting). There is no evidence that cinobufagin can take the place of digitalis.